A spin-valve-like magnetoresistance of an antiferromagnet-based tunnel junction.

A spin valve is a microelectronic device in which high- and low-resistance states are realized by using both the charge and spin of carriers. Spin-valve structures used in modern hard-drive read heads and magnetic random access memoriescomprise two ferromagnetic electrodes whose relative magnetization orientations can be switched between parallel and antiparallel configurations, yielding the desired giant or tunnelling magnetoresistance effect. Here we demonstrate more than 100% spin-valve-like signal in a NiFe/IrMn/MgO/Pt stack with an antiferromagnet on one side and a non-magnetic metal on the other side of the tunnel barrier. Ferromagneticmoments in NiFe are reversed by external fields of approximately 50 mT or less, and the exchange-spring effect of NiFe on IrMn induces rotation of antiferromagnetic moments in IrMn, which is detected by the measured tunnelling anisotropic magnetoresistance. Our work demonstrates a spintronic element whose transport characteristics are governed by an antiferromagnet. It demonstrates that sensitivity to low magnetic fields can be combined with large, spin-orbit-coupling-induced magnetotransport anisotropy using a single magnetic electrode. The antiferromagnetic tunnelling anisotropic magnetoresistance provides a means to study magnetic characteristics of antiferromagnetic films by an electronic-transport measurement.

Electrical measurement of antiferromagnetic moments in exchange-coupled IrMn/NiFe stacks.

We employ antiferromagnetic tunneling anisotropic magnetoresistance to study the behavior of antiferromagnetically ordered moments in IrMn exchange coupled to NiFe. Experiments performed by common laboratory tools for magnetization and electrical transport measurements allow us to directly link the broadening of the NiFe hysteresis loop and its shift (exchange bias) to the rotation and pinning of antiferromagnetic moments in IrMn. At higher temperatures, the broadened loops show zero shift, which correlates with the observation of fully rotating antiferromagnetic moments inside the IrMn film. The onset of exchange bias at lower temperatures is linked to a partial rotation between distinct metastable states and pinning of the IrMn antiferromagnetic moments in these states. The observation complements common pictures of exchange bias and reveals an electrically measurable memory effect in an antiferromagnet.

Rapid and simple detection of Clostridium botulinum types A and B by loop-mediated isothermal amplification.

AIMS: To develop a convenient and rapid detection method for toxigenic Clostridium botulinum types A and B using a loop-mediated isothermal amplification (LAMP) method. METHODS AND RESULTS: The LAMP primer sets for the type A or B botulinum neurotoxin gene, BoNT/A or BoNT/B, were designed. To determine the specificity of the LAMP assay, a total of 14 C. botulinum strains and 17 other Clostridium strains were tested. The assays for the BoNT/A or BoNT/B gene detected only type A or B C. botulinum strains, respectively, but not other types of C. botulinum or strains of other Clostridium species. Using purified chromosomal DNA, the sensitivity of LAMP for the BoNT/A or BoNT/B gene was 1 pg or 10 pg of DNA per assay, respectively. The assay times needed to detect 1 ng of DNA were only 23 and 22 min for types A and B, respectively. In food samples, the detection limit per reaction was one cell for type A and 10 cells for type B. CONCLUSIONS: The LAMP is a sensitive, specific and rapid detection method for C. botulinum types A and B. SIGNIFICANCE AND IMPACT OF THE STUDY: The LAMP assay would be useful for detection of C. botulinum in environmental samples.

A simple and sensitive method for detection of Bacillus anthracis by loop-mediated isothermal amplification.

AIMS: To develop a rapid and simple system for detection of Bacillus anthracis using a loop-mediated isothermal amplification (LAMP) method and determine the suitability of LAMP for rapid identification of B. anthracis infection. METHODS AND RESULTS: A specific LAMP assay targeting unique gene sequences in the bacterial chromosome and two virulence plasmids, pXO1 and pXO2, was designed. With this assay, it was possible to detect more than 10 fg of bacterial DNA per reaction and obtain results within 30-40 min under isothermal conditions at 63 degrees C. No cross-reactivity was observed among Bacillus cereus group and other Bacillus species. Furthermore, in tests using blood specimens from mice inoculated intranasally with B. anthracis spores, the sensitivity of the LAMP assay following DNA extraction methods using a Qiagen DNeasy kit or boiling protocol was examined. Samples prepared by both methods showed almost equivalent sensitivities in LAMP assay. The detection limit was 3.6 CFU per test. CONCLUSIONS: The LAMP assay is a simple, rapid and sensitive method for detecting B. anthracis. SIGNIFICANCE AND IMPACT OF THE STUDY: The LAMP assay combined with boiling extraction could be used as a simple diagnostic method for identification of B. anthracis infection.

Poor applicability of estimation method for adults to calculate unbound serum concentrations of valproic acid in epileptic neonates and infants.

OBJECTIVE: To characterize the relationship between total and unbound concentrations of valproic acid (VPA) in epileptic neonates and infants, the clinical examination records of those patients archived via therapeutic drug monitoring (TDM) activities were retrospectively analyzed. METHODS: The screening encompassed 249 records of 114 epileptic patients aged 0-19 years old, who were treated with VPA monotherapy and whose total and unbound VPA concentrations were determined. These data were divided into groups according to the patients' age. In each group, the relationship between total and unbound VPA concentrations was compared to a reference profile, and the deviation from the reference was evaluated. The reference profile was calculated using the Langmuir equation, in which two parameters Kd and Bm were set to 7.8 and 130 microg/mL, respectively, according to our previous findings. RESULTS: The relationship between total and unbound VPA concentrations of patients of 0 years old considerably deviated from the reference, and their unbound VPA concentrations were generally higher compared to the corresponding reference values. It is suggested that the large deviation is related to the fact that the serum albumin concentrations of patients younger than 1 year old tend to be lower than those of patients in other age groups. CONCLUSION: Since the relationship between the VPA concentrations of epileptic neonates and infants is noticeably different from the reference, the unbound serum VPA concentrations of these patients are not adequately estimated using the same method as that for grown-ups. The unbound VPA concentrations of neonates and infants should be explicitly determined via TDM activities.

Fifteen-year survival of resin-bonded vs full-coverage fixed dental prostheses.

PURPOSE: We clarified cumulative survival and event-free rates of resin-bonded fixed dental prostheses (RBFDPs) and compared them to those of fixed dental prostheses (FDPs) to refine risk factors for non-survival/event and use of tooth extraction after the period of non-survival/event. METHODS: Study subjects were selected among all patients who consecutively attended the Fixed Prosthodontic Clinic of Okayama University Hospital. Eligible patients were those who received 3-unit metal-framed 2-retainer (wing-wing) RBFDPs or conventional full-coverage FDPs (RBFDPs/FDPs: 129/177 prostheses). Data were analyzed by Kaplan-Meier analysis with the log-rank test, Mann-Whitney test, chi-square test, and Cox proportional hazards analysis. RESULTS: The 15-year cumulative survival rates were 66.5% for the RBFDP group and 61.6% for the FDP group, which were not significantly different (p = 0.59). The 15-year cumulative event-free rates were 53.4% for the RBFDP group and 59.2% for the FDP group, which were not significantly different (p = 0.52). No significant risk factors related to non-survival and event-free of RBFDPs/FDPs were identified in the analysis model using treatment method, sex, age, number of remaining teeth, and treatment site as explanatory variables. The number of cases in which RBFDPs/FDPs resulted in non-survival due to abutment tooth extraction was significantly lower in RBFDPs (p < 0.01). Further, the abutment tooth as a non-vital tooth was identified as a risk factor for RBFDPs/FDPs resulting in non-survival due to abutment tooth extraction. CONCLUSIONS: The present study is the first to indicate RBFDP as a prosthetic treatment option which should be selected for patients with slight or no abutment tooth decay.

Risk of hypoglycemia associated with thyroid agents is increased in patients with liver impairment.

OBJECTIVE: To determine the risk of developing hypoglycemia from drugs that affect glucose homeostasis and evaluate the elevation of that risk by liver impairment as judged by a decrease of liver reserve or the severity of abnormal values in liver function tests. METHODS: A hospital-based case-control study was carried out. The base population consisted of all patients aged 20 years and older attending a university hospital in Japan from 2002 - 2004 who had received drugs and serum glucose measurements. Cases were defined as having had at least one episode of hypoglycemia as determined by a serum glucose concentration below 70 mg/dl. Up to 5 controls for each case were matched for the year of serum glucose measurement, out- or inpatient status, clinical departments visited, and age difference within 5 years, taken randomly from the base population without hypoglycemia. The odds ratio for developing hypoglycemia was estimated using conditional logistic regression analysis. RESULTS: From a base population of 10,011, 245 cases and 1,194 controls were enrolled. Of the drugs investigated, levothyroxine use was associated with an increased risk of hypoglycemia in patients with liver impairment (adjusted odds ratio; non-use with normal liver (reference), non-use with liver impairment 0.91 (95% CI 0.62, 1.33), use with normal liver 4.50 (0.58, 34.76), use with liver impairment 14.68 (1.57, 137.4), p for trend 0.007). The risk elevation likely depended on the lowering of liver reserve. CONCLUSION: Clinicians and pharmacists should carefully monitor serum glucose concentrations in levothyroxine users with liver impairment, especially those with lower liver reserve.

Characterization of non-linear relationship between total and unbound serum concentrations of valproic acid in epileptic children.

OBJECTIVE: To establish a regression equation to properly estimate the unbound serum concentration of valproic acid (VPA) from its total serum concentration; the relationship between total and unbound serum VPA concentrations was retrospectively characterized. METHODS: Data were obtained from the clinical examination records that were routinely archived during therapeutic drug monitoring. The screening encompassed 342 records of 108 paediatric patients whose total and unbound VPA concentrations had been determined. The relationship between total and unbound VPA concentrations was characterized according to the Langmuir equation by taking account of inter-individual variability with the nonmem program. RESULTS: The total VPA concentration (C(t)) in the screened patients ranged from 5.5 to 179.8 microg/mL, and the unbound VPA concentration (C(f)) increased in a non-linear manner as the total VPA concentration increased. Taking account of the effects of antiepileptics concurrently administered, the VPA dissociation constant (K(d)) and maximum binding site concentration (B(m)) were 7.8 +/- 0.7 and 130 +/- 4.5 microg/mL respectively, for the regression equation, C(t) = C(f) + B(m) x C(f)/(K(d) + C(f)). An alteration in the unbound concentration was seen in patients who were treated with the combination of VPA and ethosuximide and in those who received two additional antiepileptics. CONCLUSIONS: A regression equation for estimation of the unbound VPA concentration, based on total VPA concentration collected during routine therapeutic drug monitoring was established. Use of two additional antiepileptics and ethosuximide treatment was considered as potential factors affecting unbound VPA concentration.

Endogenous calcitonin gene-related peptide (CGRP) mediates adrenergic-dependent vasodilation induced by nicotine in mesenteric resistance arteries of the rat.

1. The mechanisms underlying vasodilator effect of nicotine on mesenteric resistance blood vessels and the role of calcitonin gene-related peptide (CGRP)-containing (CGRPergic) vasodilator nerves were studied in the rat. 2. Mesenteric vascular beds isolated from Wistar rats were perfused with Krebs solution, and perfusion pressure was measured with a pressure transducer. 3. In preparations with intact endothelium and contracted by perfusion with Krebs solution containing methoxamine, perfusion of nicotine (1 - 100 microM) for 1 min caused a concentration-dependent vasodilator response without vasoconstriction. 4. The nicotine-induced vasodilation was markedly inhibited by hexamethonium (nicotinic cholinoceptor antagonist, 10 microM) and blocked by guanethidine (adrenergic neuron blocker, 5 microM). 5. Either denervation by cold storage (4 degrees C for 72 h) or adrenergic denervation by 6-hydroxydopamine (toxin for adrenergic neurons, 2 mM for 20 min incubation, twice) blocked the nicotine-induced vasodilation. 6. Neither endothelium removal with perfusion of sodium deoxycholate (1.80 mg ml(-1), for 30 s) nor treatment with N(omega)-nitro-L-arginine (nitric oxide synthase inhibitor, 100 microM), atropine (muscarinic cholinoceptor antagonist, 10 nM) or propranolol (beta-adrenoceptor antagonist, 100 nM) affected the nicotine-induced vasodilation. 7. In preparations without endothelium, treatment with capsaicin (depleting CGRP-containing sensory nerves, 1 microM) or human CGRP[8 - 37] (CGRP receptor antagonist, 0.5 microM) markedly inhibited the nicotine-induced vasodilation. 8. These results suggest that, in the mesenteric resistanc artery of the rat, nicotine induces vasodilation, which is independent of the function of the endothelium and is involved in activation of CGRPergic nerves. It is also suggested that nicotine stimulates presynaptic nicotinic cholinoceptors on adrenergic nerves to release adrenergic neurotransmitters, which then act on CGRPergic nerves to release endogenous CGRP from the nerve.

Regional variation in oral mucosal drug permeability.

The purpose of this article is to review regional variation in oral mucosal drug permeability. The structure and composition of the mucosa at different sites in the oral cavity, factors affecting mucosal permeability, selection of appropriate experimental systems for studying mucosal permeability, and formulation factors including penetration enhancement relevant to the design of systems for oral mucosal delivery are reviewed.

Drug elimination function of rat small intestine: metabolism and intraluminal excretion.

The metabolic and excretory function of the small intestine was investigated after oral and intravenous administration of drugs having an aromatic amino group to rats. After administration of drugs into the intestinal loop at the initial concentration of 0.1 mM, significant excretion of their N-acetylated forms into the lumen was observed. The amount of N-acetyl forms excreted in the lumen were 39.3 +/- 3.5, 63.5 +/- 20.9 and 18.0 +/- 13.8% of disappeared drugs from the lumen for p-aminobenzoic acid (PABA), p-aminosalicylic acid and sulfanilic acid, respectively. The excretion of p-acetamidobenzoic acid (Ac-PABA) after the absorption of PABA was reduced by the coadministration with salicylic acid, benzoic acid and 2,4-dinitrophenol. Salicylic acid noncompetitively inhibited the acetylation of PABA by the intestinal N-acetyltransferase. A good correlation was found between the intestinal N-acetyltransferase activities for drugs and the intraluminal excretion of N-acetyl derivatives after intestinal absorption of drugs. These results indicate that a drug having a higher susceptibility to intestinal N-acetyltransferase would undergo a greater excretion into the lumen in its N-acetyl form after intestinal absorption. After intravenous administration of PABA at a dose of 100 mumole/kg, 4.02 +/- 0.51% of dose was excreted in the lumen as Ac-PABA in 30 min. On the other hand, a significantly smaller fraction (2.72 +/- 0.68% of dose) was excreted in the lumen after intravenous injection of 100 mumole/kg of Ac-PABA. The larger excretion of Ac-PABA after administration of PABA indicates the contribution of intestinal metabolism on the transfer of PABA not only after oral, but also after intravenous administration.

Ultrasonographic and serologic abnormalities in Schistosoma japonicum infection in Leyte, the Philippines.

We have identified specific ultrasonographic changes in Schistosoma japonicum-infected patients associated with serologic indicators of general liver function. An ultrasonographic examination concomitant with hematologic and biochemical serum analyses was performed on 102 patients at the Schistosomiasis Hospital in Leyte, The Philippines. The ultrasonographic liver images were classified into four patterns, according to the development of periportal fibrosis and the patterns of echogenic bands. Eleven cases with a long-term infection showed typical septal formation (network pattern). Other ultrasonographic changes in the portal system, such as the severity of splenomegaly, did not correlate with the age of the study patients or the duration of their infection; however, the production of collateral vessels was clear in the group of older patients. Among various hematologic and biochemical serum indicators of liver damage, the serum levels of total bile acid (TBA) and procollagen-III-peptide (P-III-P) strongly correlated with the development of hepatic fibrosis and protal hypertension. These findings suggest that the ultrasonographic liver patterns classified here, along with the changes in serum levels of TBA and P-III-P, provide useful indicators for field monitoring of S. japonicum infection.

Improvement of ultrasonographic and serologic changes in Schistosoma japonicum-infected patients after treatment with praziquantel.

We previously reported ultrasonographic and serologic abnormalities in 102 patients infected with Schistosoma japonicum in Leyte, The Philippines. These patients were subsequently treated with praziquantel (3 x 20 mg/kg), and changes in ultrasonographic images and the serum levels of liver function markers in 52 patients were followed up every three months for a period of 17 months. Improvement in the thickening of the portal vein wall and the intensity of echogenic bands was detected six months after treatment with praziquantel. The level of splenomegaly was also reduced in 42 patients who originally did not show the production of collateral vessels. A significant decrease in the serum total bile acid (TBA) level was detected in all patients six months after treatment with praziquantel. However, significant ultrasonographic changes could not be detected in the patients classified as type 3, with severe hepatic fibrosis caused by the long-term infection. These results clearly show that ultrasonographic examination, along with data on the serum TBA level, provides a sensitive tool to monitor the severity of hepatic fibrosis and portal hypertension caused by S. japonicum infection, as well as the improvement resulting from praziquantel treatment.

Decreased depressor response mediated by calcitonin gene-related peptide (CGRP)-containing vasodilator nerves to spinal cord stimulation and levels of CGRP mRNA of the dorsal root ganglia in spontaneously hypertensive rats.

The depressor response to electrical stimulation of the spinal cord and the level of calcitonin gene-related peptide (CGRP) mRNA in the dorsal root ganglion (DRG) in the spontaneously hypertensive rat (SHR) was compared with the normotensive Wistar Kyoto rat (WKY) and Wistar rat (WR). The animals were pithed by inserting a stainless-steel rod into the spinal cord. Pithed rats were treated with hexamethonium (2 mg/kg/min i.v.) to block autonomic outflow, and mean arterial blood pressure (MBP) was maintained at approximately 100 mmHg with continuous infusion of methoxamine (10 to 15 microg/kg/min i.v.). Electrical stimulation (2 and 4 Hz for 30 s) of the lower thoracic spinal cord (T9-12) via the pithing rod caused a frequency-dependent depressor response without a change in heart rate. The depressor response to spinal cord stimulation was significantly smaller in SHR than in WKY and WR. Long-term treatment of 8 week-old SHR with captopril (0.1% in drinking water) for 7 weeks restored the reduced depressor response to spinal cord stimulation. The level of CGRP mRNA in DRG of SHR was significantly lower than that in WKY. These results suggest that the function of CGRP-containing nerves from the spinal cord decreases in SHR and captopril treatment prevents its reduction.

Adrenomedullin inhibits neurotransmission of calcitonin gene-related peptide (CGRP)-containing vasodilator nerves in rat mesenteric resistance arteries.

We have reported that the rat mesenteric resistance artery has innervation of calcitonin gene-related peptide (CGRP)-containing vasodilator nerves (CGRPergic nerves). We also demonstrated that adrenomedullin (AM) causes mesenteric vasodilation through activation of CGRP receptors. The present study was designed to examine the effect of AM on neurotransmission of CGRPergic nerves in rat mesenteric arteries. In preconstricted preparations without endothelium, periarterial nerve stimulation (PNS, 1 and 2 Hz) induced a frequency-dependent vasodilation. A bolus injection of CGRP (10 pmol) into the perfusate also caused a vasodilation. AM (0.1 to 10 nM) concentration-dependently caused 40% to 60% inhibition of the PNS-induced vasodilation, but AM did not attenuate vasodilation induced by exogenous CGRP injection. The inhibitory effect of AM (10 nM) on PNS-induced vasodilation was further potentiated by CGRP [8-37] (CGRP receptor antagonist, 50 nM), which attenuated the vasodilator response to the CGRP injection. Combined perfusion of AM [22-52] (AM receptor antagonist, 10 to 100 nM) resulted in further inhibition of PNS-induced neurogenic vasodilation without affecting the vasodilator response to the CGRP injection. CGRP [8-37] but not AM [22-52] antagonized vasodilation induced by AM perfusion. These findings suggest that AM presynaptically inhibits neurotransmission of CGRPergic nerves, probably decreasing CGRP release, via receptors different from CGRP receptors.

Triphasic vascular responses to bradykinin in the mesenteric resistance artery of the rat.

The vascular effects of bradykinin were studied in rat perfused mesenteric vascular beds with active tone. Bolus injections of bradykinin (1-1000 pmol) but not des-Arg(9)-bradykinin (bradykinin B(1) receptor agonist) induced triphasic vascular responses: the initial sharp vasodilation followed by transient vasoconstriction and subsequent gradual vasodilation. The triphasic vascular responses to bradykinin were abolished by FR 172357 (3-bromo-8-[2,6-dichloro-3-[N-[(E)-4-(N,N-dimethylcarbamoyl) cinnamidoacetyl]-N-methylamino]benzyloxy]-2-metylimidazo[1,2-a]pyridine) (bradykinin B(2) receptor antagonist, 0.1 microM). Endothelium removal with sodium deoxycholate and N(w)-nitro-L-arginine (300 microM) abolished the bradykinin-induced initial sharp vasodilation. Indomethacin (0.5 microM) and seratrodast (thromboxane A(2) receptor antagonist, 0.5 and 5 microM) abolished the bradykinin-induced second vasoconstriction. The bradykinin-induced third vasodilation was abolished by capsaicin (1 microM) and calcitonin gene-related peptide (CGRP)-(8-37) (CGRP receptor antagonist, 0.5 microM). These findings suggest that the bradykinin-induced initial sharp vasodilation is endothelium dependent, that endogenous thromboxane A(2) is involved in the second vasoconstriction, and that the third slow vasodilation is produced by activation of capsaicin-sensitive CGRP-containing nerves.

Adrenergic nerves mediate acetylcholine-induced endothelium-independent vasodilation in the rat mesenteric resistance artery.

Mechanisms underlying acetylcholine-induced endothelium-independent vasodilation were studied in the rat mesenteric vascular bed isolated from Wistar rats. In preparations without endothelium, and contracted by perfusion with Krebs solution containing methoxamine (2-7 microM), perfusion of acetylcholine (1-100 microM) for 1 min produced a concentration-dependent vasodilation. Denervation of denuded preparations by cold storage (4 degrees C for 72 h) abolished the acetylcholine-induced vasodilation; 10 and 100 nM atropine abolished 1 and 10 microM acetylcholine-induced vasodilation, but it inhibited only 20% of vasodilation by 100 microM acetylcholine. The acetylcholine-induced atropine-resistant vasodilation was inhibited by 10 and 100 microM hexamethonium, 5 microM guanethidine, 50 microM bretylium, in vitro 6-hydroxydopamine (2 mM for 20 min, twice), 1 microM capsaicin and 0.5 microM calcitonin gene-related peptide (CGRP)-(8-37) (CGRP receptor antagonist). These findings suggest that the acetylcholine-induced endothelium-independent nicotinic vasodilation requires the presence of intact adrenergic nerves, and is mediated by endogenous CGRP released from CGRP-containing nerves.

Malleus bar as a rare cause of congenital malleus fixation: CT demonstration.

Among congenital ossicular anomalies without external ear atresia, malleus fixation is least common. We report a case of congenital malleus fixation by a bony bar connecting the malleus neck to the posterior tympanic wall, which was depicted on thin-section CT scans.

Biopharmaceutical evaluation of salicylazosulfanilic acid as a novel colon-targeted prodrug of 5-aminosalicylic acid.

A prodrug of 5-aminosalicylic acid (5-ASA), salicylazosulfanilic acid (SASA), which consists of sulfanilic acid linked to 5-ASA through an azo-linkage was newly synthesized. Biopharmaceutical properties of SASA were evaluated in comparison with those of salicylazosulfapyridine (SASP) in rats. Since SASA is much more hydrophilic than SASP, the absorption of SASA from the small intestine was less in comparison with SASP. When SASA and SASP were incubated with the rat intestinal contents under anaerobic conditions, both compounds were stable in the small-intestinal contents, but were rapidly degraded to 5-ASA in the cecal and the colonic contents. The degradation to 5-ASA by the large-intestinal contents was suppressed by the pretreatment with kanamycin sulfate, suggesting that the bioconversion of SASA is mediated by the intestinal microflora similarly to that of SASP and that SASA is also a prodrug of 5-ASA. After the oral administration, 5-ASA was found neither in the stomach nor in the small intestine in case of both prodrugs. Most of the prodrugs were transferred to the lower intestine where they were degraded to 5-ASA. The recovery of SASA including the metabolites from the gastrointestinal tract at four hours after the oral administration was significantly greater than that of SASP. Accordingly, SASA is free from the liberation of sulfapyridine, the adverse effect moiety of SASP, and less absorbable in the small intestine. Thus, the beneficial characteristics of SASA as an excellent colon-targeted prodrug of 5-ASA were clarified.

Perfusion cells for studying regional variation in oral mucosal permeability in humans. 2. A specialized transport mechanism in D-glucose absorption exists in dorsum of tongue.

To clarify the site of d-glucose absorption in human oral cavity, newly designed perfusion cells were applied to five different sites in the human oral cavity, i.e., the dorsum of the tongue, the ventral surface of the tongue, the labial mucosa, the floor of the mouth, and the buccal mucosa. The solution of D-glucose was perfused for 1 h and the rate of absorption was calculated from the amount that disappeared from the perfusate. D-Glucose was absorbed rapidly from the dorsum of the tongue and the absorption was saturable. The saturable absorption was also observed in the ventral surface of the tongue, but not in the other three sites. The rate of D-glucose absorption in the dorsum and the ventral surface of the tongue was significantly larger than that of L-glucose, while in the other sites they were not significantly different. The presence of a specialized transport system for D-glucose absorption in the dorsum of human tongue was suggested.